How big can the U.S. current account stay?

In the past few years, the U.S. current account deficit has shrunk from over 6% of GDP in mid-2006 to less than 3% today. Since these current account deficits reflect capital account surpluses, many people view them as a symptom of the problems that led to the crisis. That is, funds from abroad were fueling the credit boom in the United States, which in turn fed the boom in housing prices, etc.

As you can see in the chart below, over the past three decades the U.S current account has been in surplus only briefly in the first half of 1990. Since then, it has been continuously in deficit. How is it that the United States can keep borrowing without a collapse in the currency or a surge in borrowing costs?  Is there some sort of limit? 

One possible answer is that because the world runs on U.S. dollars, everyone needs U.S. dollar-denominated securities. Countries use these both to transact and to insure themselves against foreigners’ suddenly deciding to withdraw assets – a capital flow reversal. These needs result in the “exorbitant privilege” that accrues to the United States as the issuer of the reserve currency.

U.S. Current Account Deficit
(as a percentage of GDP, quarterly)

Source: OECD. Data are from FRED (BPBLTT01USQ188S) from 1980 to 1Q 2014.

The natural issuer of these dollar assets is a U.S. entity (either private or public). So, meeting this demand for U.S. dollar instruments means that – other things equal – the United States will run a capital account surplus. That, in turn, means a current account deficit.

But how big a deficit? To get a rough answer, we introduce a few definitions and some algebra. Define Y^R as nominal GDP for the countries that demand a level R of reserve currency assets (mostly the emerging market and developing economies); Y^US as the nominal GDP of the US, α* as the target ratio of nominal reserves to nominal GDP measured in the reserve currency (R/Y^R); β as the relative size of Y^US to Y^R, the ratio of the supplier to the demander’s nominal GDP (Y^US=βY^R), and g^R is the growth rate of the reserve demander’s nominal GDP.

Now we can do a bit of algebra to derive the following expression for the ratio of the change in reserves relative to the GDP of the reserve currency issuer:

This is the U.S. current account deficit that arises from the fact that countries want to hold U.S. dollar assets as reserves.

Sources:  Data on reserves is total reserves from the IMF’s International Financial Statistics (IFS), and GDP is from the IMF World Economic Outlook Database. Data is for year 2012.

Now, we need to compute three things: (1) the target ratio of U.S. dollar reserves to GDP for the reserve holding countries (α*); (2) the ratio of U.S. nominal GDP to the nominal GDP of those countries that hold reserves (β), and; (3) the nominal GDP growth rate of the reserve-holding countries (g^R). Starting with reserve holding, the IMF currently reports that countries hold roughly $11.7 trillion in foreign exchange reserves. Focusing only on emerging market and developing economies (EMDEs), reserves are currently $7.9 trillion, or 27% of their GDP. The bulk of this is in U.S. dollars, so we can conservatively set α* = 0.2. (The bar chart above plots an estimate of α* for the 25 countries holding 85% of reserves in the world in 2012.) Turning to the ratio of GDP levels, U.S. GDP is currently something like 60% of EMDE GDP. (See the figure below.) This means that β=0.6. And finally, there is nominal GDP growth in EMDEs. While this number has been between 5% and 6% in the past few years, in the past decade, the level has been over 10% per year. The IMF’s forecasts for the next few years suggest a level that is close to 7%, so we set g^R equal to 7%.

Source: International Monetary Fund.

Putting this all together, we have the following estimate of the current account deficit that the U.S. economy could sustain so long as reserve demand from the rest of the world for U.S. dollar assets remains stable:

Annual demand for U.S. dollar reserves =  2.3% of U.S. nominal GDP (as in the equation below).

Where are the possible errors in our estimates? We see three sources. First, there are the reserve holdings. By focusing on the average in the EMDEs, we obscure the potentially dominating behavior of China. China holds reserves of about $4 trillion, which is nearly 50% of Chinese GDP (at market exchange rates). If China chooses to stop buying U.S. assets, and no one else picks up the slack, then α* will fall and so will our estimate of the size of the current account that can be supported by this mechanism.

Second comes the relative size of the U.S. economy, β. Here, it is pretty clear that the United States is growing more slowly than the EMDEs, so β is shrinking. (As the chart shows, the level has fallen steadily over the past 20 years.) This will increase demand for U.S. assets as a fraction of U.S. GDP, so it would appear to raise the current account that can be financed. In the short run, that seems right. But in the long run, as the relative importance of the U.S. economy wanes, so will the desirability of holding U.S. dollars. In the first half of the 20th century, the world economy was based on British pounds. What eventually happened to the pound could also happen to the dollar.

Finally, there is the nominal growth rate of EMDEs. Here, we have been conservative. The actual growth rate of nominal GDP (g^R) could easily exceed 7%, as it has for much of the last quarter century.

Overall, these simple computations lead us to conclude that the U.S. current account deficit can remain at 2% of GDP or more for some time to come without threatening the value of the dollar or triggering an externally driven surge in U.S. borrowing costs.